LEVELS OF PROCESSING IN FACIAL MEMORY

brasscoffeeΤεχνίτη Νοημοσύνη και Ρομποτική

17 Νοε 2013 (πριν από 3 χρόνια και 4 μήνες)

51 εμφανίσεις

LEVELS OF PROCESSING IN
FACIAL MEMORY




Sarah Babcock, Rose Ann Calvieri,

Lauren Cudney,
Vedran Dzebic ,Silvia Eleftheriou, and Jeff Mazurkewich
,



TOPIC DEVELOPMENT



Progression of ideas for possible studies:




Initial thought: Combine spatial memory with decision
making task



Replaced spatial memory with word memory



Word memory is well studied, we wanted to approach
memory from a different angle



Memory involving different levels of processing
regarding objects



Decided to examine faces instead of objects


RATIONALE FOR THE STUDY




We encounter faces constantly, making facial
memory a crucial skill for social interactions




Examining techniques of facial memory can
potentially improve ease of everyday social
interactions




LITERATURE: CRAIK & TULVING
(1975)



Experiment


Processing words at different levels


Subjects presented with words & asked questions to
various levels


Shallow = same font


Medium = counting letters in word


Deep = synonyms



Results


Better memory recall for words processed at a deeper
level

L
ITERATURE
: B
RUCE

&
VALENTINE

(1985)



Experiment


Priming task: Subjects were shown the name or
photo of a celebrity


Recognition task: Subjects were then shown a series
of photos of celebrities and asked to name them


The subject had seen either the name, same photo or
a different photo of the same celebrity in the priming
task



Results


Better memory for faces which subject’s had seen
pictures of in the priming task (same or different)


Experiment


Source memory for faces (internal or external)


Questions were asked about faces in a presentation of
facial images


The subject either generated the answer or the
answer was provided for them (accompanied the face)



Results


Subjects were better at source memory for faces
which asked them to generate answers (internal),
than for faces accompanied by the answers (external)

LITERATURE: GEGHMAN & MULTHAUP
(2004)

PURPOSE OF THE EXPERIMENT



Purpose:


To determine whether different questions can elicit
deeper levels of processing



To establish if deeper levels increase subsequent
memory on a facial recognition task


LEVELS OF PROCESSING


Questions
:


1) What is this person’s most attractive feature?


2) What job do you think this person has?


3) How old is this person?


4) What is this person’s gender?




Examine whether any of these questions will
result in deeper level of processing, measured
by accuracy of facial recognition

HYPOTHESIS


If different levels of facial processing can be
achieved, deeper level processing will lead to
better recognition of faces



Hypothesized to elicit
shallower

processing:


Questions about gender and age


Hypothesized
deeper

processing:


Questions about attractiveness and occupation



Kirkland, Reynolds and Pezdek (1992)



M
ETHODS



Subjects




30 subjects


24 in experimental group


6 in control group


All Mac undergrads


Age range 18
-
24, Mean 20








M
ETHODS


Stimulus/Materials



Study Task


32 faces



Experimental group


Each Face paired with one question



Control group


No questions presented


M
ETHODS


Stimulus/Materials



Recognition Task


60 faces (28 novel)



Have you seen this face in the previous presentation?


Yes/No responses



All subjects given same task


M
ETHODS


Cover Story



Study Task


Subjects were told:


There will be questions about the faces


They need to answer as quickly as possible the questions


We are looking at how much you can tell about a person by
their appearance



Recognition Task:


Subjects were naïve of recognition task to follow the study
task




LEVELS OF PROCESSING


Questions
:


1) What is this person’s most attractive feature?


2) What job do you think this person has?


3) How old is this person?


4) What is this person’s gender?



Examine whether any of these questions will
result in deeper level of processing, measured
by accuracy of facial recognition

SLIDESHOW EXAMPLE: STUDY TASK

A
TTRACTIVENESS
?

A
GE
?

G
ENDER
?

J
OB
?

S
LIDESHOW

E
XAMPLE
: R
ECOGNITION

T
ASK


DATA COLLECTION


Study Task


Recorded subject’s responses to questions




Recognition Task


Recorded if the subject answered yes or no

STUDY TASK DATA SHEET


Subject
#

Question Orders

Picture #

Group Number

S’s
response

1

2

3

4

1

a

d

b

c



2

b

c

a

d



3

c

b

d

a



4

d

a

c

b



5

a

d

b

c



6

b

c

a

d



RECOGNITION TASK DATA SHEET


Subject #

Recognition Task Ss Response Sheet

Recognition
task slide #'s

Picture Number

1

2

3

4

S’s
response

S’s
Correct Responses

2

54

0

0

0

0





4

21

a

d

b

c





6

27

c

b

d

a





8

45

0

0

0

0





10

43

0

0

0

0





12

14

b

c

a

d





14

41

0

0

0

0





16

28

d

a

c

b





18

24

d

a

c

b





RESULTS


Group Results (Experimental & Control)



Independent t
-
test



Descriptives



One
-
way ANOVA



Post
-
hoc (Bonferroni)




GROUP STATISTICS

Group Statistics

Group

N

Mean

Std. Deviation

Std. Error Mean

Sstotal

1

24

49.4167

5.19964

1.06137

2

6

40.6667

6.37704

2.60342

INDEPENDENT T
-
TEST

Independent Samples Test

Levene's

Test for

Equality of

Variances

t
-
test for Equality of Means

F

Sig.

t

df

Sig.

(2
-
tailed)

Mean

Difference

Std. Error

Difference

95% Confidence

Interval of the

Difference

Lower

Upper

Ss

total

Equal
variances
assumed

.033

.858

3.531

28

.001

8.75000

2.47783

3.67440

13.82560

Equal
variances
not
assumed

3.112

6.760

.018

8.75000

2.81146

2.05373

15.44627

DESCRIPTIVES

Descriptives

Data

N

Mean

Std. Deviation

Std. Error

95% Confidence Interval for
Mean

Minimum

Maximum

Lower Bound

Upper Bound

1

24

6.7917

1.02062

.20833

6.3607

7.2226

5.00

8.00

2

24

6.0833

1.44212

.29437

5.4744

6.6923

3.00

8.00

3

24

5.2500

1.89393

.38660

4.4503

6.0497

1.00

8.00

4

24

5.7500

1.32698

.27087

5.1897

6.3103

3.00

8.00

Total

96

5.9688

1.53865

.15704

5.6570

6.2805

1.00

8.00

ONE WAY ANOVA

ANOVA

Data

Sum of
Squares

df

Mean
Square

F

Sig.

Between
Groups

30.115

3

10.038

4.741

.004

Within Groups

194.792

92

2.117

Total

224.906

95

POST HOC

Multiple Comparisons

Data

Bonferroni

(I)

Questions

(J)

Questions

Mean Difference
(I
-
J)

Std. Error

Sig.

95% Confidence Interval

Lower Bound

Upper Bound

1

2

.70833

.42005

.571

-
.4243

1.8410

3

1.54167
*

.42005

.002

.4090

2.6743

4

1.04167

.42005

.090

-
.0910

2.1743

2

1

-
.70833

.42005

.571

-
1.8410

.4243

3

.83333

.42005

.301

-
.2993

1.9660

4

.33333

.42005

1.000

-
.7993

1.4660

3

1

-
1.54167
*

.42005

.002

-
2.6743

-
.4090

2

-
.83333

.42005

.301

-
1.9660

.2993

4

-
.50000

.42005

1.000

-
1.6327

.6327

4

1

-
1.04167

.42005

.090

-
2.1743

.0910

2

-
.33333

.42005

1.000

-
1.4660

.7993

3

.50000

.42005

1.000

-
.6327

1.6327

*. The mean difference is significant at the 0.05 level.

LIMITATIONS OF CURRENT STUDY


1) Response time:



5 second time limit


Subjects provided arbitrary responses


May have focused more on the question than actual
face


Further studies: allow slightly more time for subjects’
responses

LIMITATIONS OF CURRENT STUDY


2) Occupation Question:



Was predicted to elicit deeper processing



Answer didn’t require face processing


Future studies: questions relying more on facial
features


(I.e ethnicity, cosmetic surgery)


LIMITATIONS OF CURRENT STUDY


3) Facial Images:



Atypical facial images compared to participants and
people within the participants’ environment


Non significant results maybe due to quality of faces



Faces presented in grey
-
scale : require deeper
processing


Future studies: Use of coloured images is more realistic


Use more updated faces similar to those within
participants’ social environment

LIMITATIONS OF CURRENT STUDY


4) Number of Images:



Study task: 32 faces


Recognition task: 60 faces


Future studies: More faces in study and recognition
tasks


Increase power

LIMITATIONS OF CURRENT STUDY



5) Control Group:



Smaller than experimental group



Future studies:


Larger number of subjects in control group size =
increase power


Within
-
subjects control group:


Have some pictures without any questions in the
slideshow

I
MPLICATIONS


If significant:




Faces and words are processed similarly



Improve peoples ability to remember new
acquaintances



Eyewitness testimony


CONNECTIONS TO PREVIOUS
RESEARCH



Verbalization and conceptualization of faces lead to
better facial recognition


Itoh, 2005; Bruce & Valentine, 1985



Improved memory when face is paired with question,
and when the answers are generated


Geghman & Multhaup, 2004



Levels of processing may have had an effect on facial
recognition


Craik, & Tulving, 1975




Word memory vs. Facial memory



Mechanisms by which words are processed may not
be the same mechanisms employed in facial
recognition



Similar processing may be involved


CONNECTIONS TO PREVIOUS
RESEARCH

FUTURE RESEARCH




Intentional vs. Incidental


Previous works show there is no difference in
memory if the learning is incidental or intentional


Craik, & Tulving, 1975



Examine if intentional or incidental learning has an
effect on facial recognition




FUTURE RESEARCH



Facial images


Examine the recognition of faces more typically seen
in the subject’s environment



Investigating recognition of faces of varying
ethnicities



Neuroimaging: fMRI


Examine areas of activation between different
questions


Compare word processing to facial processing